aveiga_2014_num2

aveiga_2014

Alvaro Veiga (visiting Professor FIME - Pontifical Catholic University of Rio de Janeiro)

In this talk we present the main facts about the Brazilian energy market and some statistical tools designed to help solving investment issues concerning renewable energy. First, we describe a framework for modeling and simulating scenarios of renewable energy production coherently with energy spot prices. For this purpose, we present a monthly multivariate stochastic model for wind and hydro inflow as well as an efficient maximum likelihood estimation method for high-dimensional data based on LASSO. Secondly, we present some simple models to tackle lack of wind data in Brazil, allowing reconstructing past historical from satellite meteorological data and nearby measurement stations.